Project Summary/Abstract Disturbances in thinking are a fundamental component of the impairments in cognition associated with serious mental illnesses. The fact that compromised cognitive function transcends formal diagnostic categories makes these disturbances especially relevant to the RDoC domain of cognition. The thought disorder phenotype is of special interest, not only because of its non-specific association with serious mental illness, but also because of its exceptionally high recurrence among clinically unaffected first-degree relatives of probands. Although thought disorder per se is non-specific for any single diagnostic syndrome, certain thought disorder profiles are preferentially associated with specific disorders (e.g., schizophrenia and bipolar disorder). In this application, we propose to examine whether distinct qualitative and quantitative features of thought disorder are associated with each illness and also selectively aggregate among clinically unaffected first-degree relatives with odds ratios and effect sizes that would support their utility in genetic studies. Preliminary data support the convergent and discriminant validity of different thought disorder phenotypes in bipolar and schizophrenia probands and their respective first-degree relatives. Further, recurrence of these thought disorder phenotypes in clinically unaffected relatives is much higher than either clinical disorder among family members. Having thought disorder phenotypes for two disorders subsumed under the category of serious mental illness would be extremely useful in genetic studies, because it would allow probands and relatives to be combined independent of specific diagnosis. Moreover, the combination of the two thought disorder phenotypes is a more powerful discriminator of relatives of probands with serious mental illness than either is alone. The sample will be comprised primarily of probands with a diagnosis of bipolar disorder with psychotic features, their first- degree biological relatives and nonpsychiatric controls. A large dataset on probands with a diagnosis of schizophrenia and their first-degree relatives already exists and will be pooled with the newly collected data. The results will significantly enhance the identification of non-penetrant gene carriers among clinically unaffected relatives of probands with serious mental illnesses (SMI) and set the stage for using the composite SMI TD phenotype as a cognitive biomarker in future genetic studies.